International Conference On Applied Economics – ICOAE 2010 641
MOVING IN – SELLING OUT: THE OUTCOMES OF SLUM REHABILITATION IN MUMBAI
PAULA RESTREPO1
Abstract
Slums have been present in developing cities landscapes for quite some time and while central and local governments continue their
quest for making their cities slum-free, the overall outcome of slum policies is in most cases unknown or unclear. Some studies have
revealed that slum policies might have unforeseen side-effects such as poverty recycling: where poor slum dwellers aren't able to
maintain new living conditions and sell/rent their dwelling moving back to the slum. This paper focuses on the effects of slum
rehabilitation on residential mobility in order to test the myth of poverty recycling. In order to do so, an exhaustive household survey
was carried out in the city of Mumbai, comprising 510 households in 4 rehabilitated sites and 5 to-be-rehabilitated sites. Results show
that the magnitude of poverty recycling is small and that in most cases slum rehabilitation actually serves as a platform to attain better
living conditions both for those who left as for new comers.
JEL codes: R200 0120 D100
Key Words: slums, housing policies, Mumbai
1 Introduction
70 millions inhabitants are added each year to cities worldwide or the equivalent of creating 7 new mega-cities from scratch. And
while most of this urban growth is occurring in developing countries, about half of the new population is being absorbed by the
informal housing sector. By 2030-2040, estimates suggest that there will be 2 billion slum dwellers and slums will be housing about
one third of the total urban population (United Nations, 2007). The recognition of slums as a menace to the ―planned city‖ along with
their high correlation with poverty have made slum improvement a priority for many local governments and international
organizations. Slum and housing policies have multiplied, bringing new waves of thinking and original solutions that consider much
more than just a house. Furthermore, recent studies have revealed that slum policies, by changing the living environment, might
trigger other important positive changes as well as some unwanted side-effects. Among the stated benefits of slum policies are
poverty reduction, higher gender equality, better access to credit and higher income generation (Aiga and Umenai, 2002; Cattaneo et
al., 2009; Field, 2007).
This paper will focus on one of the possible side effects of slum rehabilitation policies: policy induced residential mobility. The
idea that many of the slum dwellers move out after policies has constantly drawn the attention of both policy makers and researchers.
On one hand, policies like titling or slum upgrading can be generators of higher mobility of households, who now can sell or rent
their dwelling in the formal market. On the other hand, slum policies might be generators of forced mobility as some slum dwellers
might be unable to face additional financial liabilities (taxes, maintenance, legal electricity) or have mismatching preferences with
new living conditions (economic activities or household size). The latter might be forced to sell/rent and move back to slums. The
Slum Rehabilitation Scheme (SRS), main slum policy of the city of Mumbai, will serve to illustrate the residential mobility analysis.
This megapolis, despite being the economic capital of India, has half of its population living in slums and introduced an innovative
slum policy based on a market approach.
The analysis of residential mobility is important in a number of ways. First, because without it the results of policy impact
measurements are generally biased. If some people have moved out and others have replaced them, the measurement of policy effects
will be bias as the study group will be composed of real beneficiaries and new comers. Second, because by studying residential
mobility it is possible to have insight of what is working and what is not. If slum dwellers are still unable to afford and sustain
themselves in the new living conditions, informality might be linked to both to an access (entering formality) problem as a
sustainability issue (staying in formality). Finally, residential mobility analysis is essential to measure accurately the policy effect in
terms of net slum-absorption and policy completion. If mobility is associated with high rates of poverty recycling the policy is simply
displacing the slum.
The main question in this paper is why some households move out of the rehabilitated buildings while others manage to stay and
what are the overall effects of the policy - considering residential mobility - in terms of net slum resorption. The analysis will be
based on a household survey carried out by the author, comprising 510 household in 9 slum pockets. The rest of the paper is
organized in the following way. The second section reviews residential mobility theories adapted to low income households and slum
settlements. The third section gives a description of the main features of the Slum Rehabilitation Scheme, a brief outline of the
methodology used in the survey and shows some evidence of residential mobility post-rehabilitation. The fourth section analyzes the
determinants of residential mobility and tests the thesis of poverty recycling based on household survey results. Finally, in the fifth
section conclusions are outlined.
2 Low-income household’s residential mobility
Residential mobility is a common subject present in economic journals but most of the literature deals with residential mobility in
developed countries and concerns only the formal housing sector. There are however a series of articles (Edwards, 1983,Gilbert and
Ward, 1982, Gilbert, 1999) in the 80s in which mobility of poor households was studied. Most of these studies sought to test a series
of hypothesis established earlier by Turner (1968), which suggested a common residential mobility pattern behind poor migrant‘s
urbanization. According to Turner, low-income household's choice was a function of three variables: (1) tenure or the choice between
renting and owning; (2) location or the proximity to employment opportunities; and (3) shelter or an individual preference of having
1 phD candidate (economics), CERNA - Mines ParisTech; [email protected] ; (33)140519042
642 International Conference On Applied Economics – ICOAE 2010
modern standard housing. His observations suggested that poor migrants (bridgeheaders) initially had low preferences for ownership
and opted for cheap rental accommodation but gradually integrated the city becoming consolidators.
The simplicity of the two-stage Turner model was latter contradicted by a series of empirical studies that evoked that to fully
understand low income household's residential mobility, both preferences and constraints needed to be taken into account(Gilbert and
Ward, 1982). Poor households might have high preferences for ownership but remain unable to afford buying a house. These studies
also suggested that the living environment played a fundamental role in the moving decision. According to Edwards (1983)
―residential mobility among poor families is as much a reaction to changing conditions in the housing market as a response to
variations in household demand‖.
Leaving aside the effects of the living environment on residential mobility, the analysis of the determinants of ―moving out‖ can
also be done with the use of housing choice theory. As (McFadden and others, 1978) explain, household's utility when choosing a
given housing solution depends not only on the housing attributes but on household's characteristics. Under this scope, residential
mobility cannot be described by a general trend of low-income migrants which are identical in preferences and constraints but as an
individual response of households to changes in their preferences (utility) and income (or budget constraint). Household‘s mobility
might be associated to changes in family size or structure, or be the product of new financial circumstances. Mobility might be
avoided - in some cases - by adapting housing attributes to preferences, as done in many Latin America cities in which low income
households build additional floors. However, attributes adaptation is in most cases restrained by space, housing structure or local
municipalities. For instance, improving household connection to water services needs higher levels of community cooperation or
political initiatives, and can rarely be the product of an individual household decision.
Residential mobility in informal settlements, excluding the cases of forced eviction, is generally restricted due to incomplete
tenure systems. As slum dwellers don't have traditional rights over their houses, selling or renting is risky and can only be achieved
through common agreements (De Soto, 1990). Lall et al. (2006) provides evidence, from slums in Bhopal that suggest that even when
slum dwellers have attained a certain security of tenure they seem to remain incapable to capitalize the value of their dwelling and
move to formal housing. Gilbert (1999), found that residential mobility of owners in consolidated low-income settlement in the city
of Bogota (Colombia) is also extremely limited. Furthermore, recent empirical studies point out that when slum housing markets
exist there seems to be very restricted vertical integration with traditional housing markets (slum to formal and formal to slum
mobility) (Gulyani and Talukdar (2008), Lall et al. (2006)).
Though there is little quantitative information of mobility changes following slum policies there does seems to be a great amount
of anecdotal evidence in the literature. Generally two causes are mentioned to be at the core of residential mobility following slum
policies: (1) increased values of property might generate a gentrification of the rehabilitated area pushing the poor out, and (2) the
entry to the formal style of living imposes unaffordable costs to households that end up leaving (poverty recycling). To the moment,
three studies, in Delhi, Cairo and Cape Town, suggest that around 10-30% of the slum original occupants move out after the slum
policy (Payne, 1977; Barry, 2006; Jacobsen, 2003). And while this gives the idea that residential mobility after slum policies might
be considerable, there is little evidence to support the theory of poverty recycling since the type and location of destinations of those
who left remains unknown.
3 Moving in
3.1 Slum Rehabilitation Scheme
Slums have made a part of the city of Mumbai for a long time. They emerged in the mid-nineteenth century and by the time of
India's independence the city had already 5% of the population housed in this type of habitat. Since then, slums have grown
considerably both in absolute and in relative terms. The total slum population passed from 2.8 millions in 1976 to 6.2 millions by the
year 2000 (MCGM, 2005). The latest report, estimated that 55% of the city population lived in informal settlements while occupying
only 16% of the city land (Hagn, 2006), a clear evidence of overcrowding and spatial inequalities of this megapolis. Living
conditions in Mumbai slums are variable but most of the settlements are relatively old and have achieved a certain degree of
consolidation. In terms of surface, slum dwellings are quite small, with 42% having 10 sq.mts, 38% between 15-20 sq.mts and only
9% above 20 sq.mts (Montgomery and Yuva, 2001). Most houses are constructed with pucca2 materials but the provision of basic
services varies considerably in each zone.
The current Slum Rehabilitation Scheme is the product of years of ―learning by doing‖ and a result of the evolution of slum
policies in the city of Mumbai. In 1976, the first census of slums was done and in 1983 a task force was created to discuss housing
and urban development issues. Despite the recognition of slums, the predominant policy in the 70's was forced demolition and
clearing of slum settlements. This policy has, unfortunately, not been completely eradicated from the city. From 1985 to 1991 three
different slum policies were implemented: the Slum Upgrading Program (1985-1991), the Prime Minister's Grant Project (1985-
1991) and the Slum Redevelopment Scheme (1991-1995). However, none of them achieved significant results in terms of slum
absorption (Mukhija, 2001).
As a result a new policy was introduced in 1995 called the Slum Rehabilitation Scheme (SRS) to substitute the Slum
Redevelopment Scheme. The SRS created a better mechanism for cross-subsidizing slum projects using two types of incentives:
Additional Development Rights (ADR) and Transfer Development Rights (TDR). It works in the following way. First, a builder or
developer associates with a slum community and collects signatures of agreement of at least 70% of the eligible slum dwellers.
Second, the project needs to follow a set of administrative procedures at the Slum Rehabilitation Authority (SRA) which is a
centralized agency created to manage the SRS. Once the project is approved, slum dwellers are relocated in transit camps, the slum is
demolished and new buildings are constructed following a standard procedure. As construction finishes slum cooperatives are formed
2 A pucca structure is one having walls and roofs made of pucca materials. Cement, burnt bricks, hollow cement/ash, bricks, stone; etc constitute the
list of pucca materials. NSS Report 486 _Condition of Urban Slums: salient features_
International Conference On Applied Economics – ICOAE 2010 643
and tenements are allotted. Slum dwellers get free housing, with basic amenities and legal titles. The builder is compensated in two
ways. If there is enough space to build additional housing units, the constructor is granted ADR. These allow the constructor to
exceed standard Floor Space Index (FSI) regulations, constructing additional housing units in the same site which he can sell on the
free market gaining profits. For example, if the project is located in the suburbs for every FSI use for rehabilitation 1 FSI in form of
ADR is granted. There is however a maximum level of FSI that cannot be exceeded which brings us to the second form of
compensation. If there is not enough space to use all ADR, the constructor is granted TDR which he can sell on the market and other
constructor can use to build additional space in another project in the city3.
3.2 Evidences of mobility following slum rehabilitation
In order to analyze household's mobility in Mumbai, three different sources of information will be taken into account. The first
two, carried out by the Tata Institute of Social Sciences, will be used as a reference for comparison of abandonment rates while the
third, carried out by the author, will be the base for the analysis of mobility determinants and abandonment destinations. The survey
carried out by the author in cooperation with the Slum Rehabilitation Society comprises a sample of 510 households spread over 9
slum pockets. This survey was carried out to fill-in the blanks of the previous two surveys and answer a series of existing questions
present in slum studies literature. Apart from residential mobility, issues like time allocation, education, access to credit and basic
services provision were taken into account. Since the two available studies only had household data post-rehabilitation, two types of
settlements where sampled to have a benchmark comparison. The first - referred as the treated group - corresponds to 4 slum pockets
that have already benefited from the SRS policy, the second - referred as the control group - corresponds to 5 slum pockets that are in
the process of being rehabilitated. At the time of the survey most of control group settlements had already collected 70% or more of
the required signatures and passed administrative procedures at the SRA to prove eligibility of slum dwellers. To assure equivalence
between treated and control groups, sites were chosen from a universe of settlements based on a series of characteristics (household
structure, area, tenure and basic services). The final group of settlements selected was very similar: around 60% gained area after
rehabilitation, most of the slum houses structures were made of durable materials and a very small proportion had separate
bathrooms, inside piped water connection or toilets. For the purpose of this article, households that didn't belong to the original group
of beneficiaries will be referred as new comers. Households that benefited from the initial policy will be referred as original
occupants.
New comers were differentiated from original occupants using a test questionnaire where households were asked a series of
questions on their previous and current living conditions. When the test questionnaire result was positive, meaning that the household
belonged to the new comers group, the interviewer moved to the new comers module to ask the occupants additional questions about
their tenure status, their previous living conditions and relationship to original occupants. When the test questionnaire results were
negative, interviewers moved to the following modules where more questions about the initial slum settlement were asked (basic
services, type of housing, etc). At this point it was easy to identify if the test questionnaire had given a false negative, since most of
the new comers were unable to respond to simple questions about their previous living status in the slum, and move to the new
comers module. Before the real survey took place, 50 questionnaires where done in a settlement - different from the 9 slum pockets -
to evaluate if the test and questions were actually working and assure that interviewers understood the methodology.
Table 1 describes sample sized and methodologies along with results for average abandonment rates and occupancy tenure status
found in the three surveys. Differences between abandonment rates (% of households that moved out) in each of the surveys might be
explained by differences in the samples selection, survey methodologies and projects characteristics. In the MUTP project, slum
dwellers were relocated to distant sites while two other surveys analyzed in-situ rehabilitation. Relocating slum dwellers in distant
areas might have serious consequences in their incomes and increase considerable transportation costs, both of which might have
effects on mobility.
Apart from differences in survey methodology, general observations suggest that the samples selected for each of the studies are
quite different. The sample selected by the author was taken from slums that had achieved a high degree of consolidation - most of
which had pucca structures - while only 20% of the SRA - TISS survey had pucca houses. Changes in the living environment, such
as evolution in the housing market can also be source of divergence but it is not possible to account for this effect. In light of the
above, abandonment rates presented in table 1 should not taken for comparisons but as an evidence of the phenomena that is taking
place.
Table 1: Abandonment rates
Survey Sample size (hhs) Projects
surveyed
Abandonment rate(%
hhs that moved out)
SRA-TISS (2003) 2138 (10% per
project)
151 13.5%
MUTP – TISS
(2008)
1505 (20% per
project)
3 15.2%
RESTREPO
(2009)
510 (35% per
project)
4 9.2%
The SRA-TISS (2003) survey included two questions that asked policy beneficiaries how many times they had moved before
rehabilitation and reasons for moving. Results indicate that for 78% of the policy beneficiaries, slum rehabilitation represented their
first shift in dwelling which suggest that residential mobility before rehabilitation took place was very low and confirms existent
literature.
3 For more information on SRS incentive mechanisms please refer to Chandy, 2007
644 International Conference On Applied Economics – ICOAE 2010
4 Selling out
―For a poor squatter in the middle of the capital city, the promise of a title would seem to be a road to riches. In practice, it is
more like a sign taped to his back that says, ―kick me‖. John Gravois - The De Soto Delusion
4.1 The determinants of moving out
The literature of residential mobility explains how households decide to move following changes in their individual preferences,
constraints or due to evolutions in the living environment. Therefore, to determine mobility causes following SRS in Mumbai, each
of the aspects of the household choice equation will be examined. If the hypothesis of a constant household utility is held, or that
household preferences present little or no change when passing from the slum household to rehab tenements, the analysis of
residential mobility can lay on changes in housing attributes and evolutions of household‘s budget constraint. A broader approach,
considering changes in the living environment like evolutions in the housing and job market is unfortunately beyond our capacity.
One could argument that changes in the living environment affect all households equally but this affirmation is not exact since
households living in the same area might be affected differently by the global crisis or the relocation of industrial companies in the
periphery of the city. This lack of information will remain one of the downfalls of the analysis.
4.2 Changes in housing attributes
The Slum Rehabilitation Scheme produces radical changes in some of the housing attributes. As table 2 shows, SRS provides
pucca structured houses, legal titles and individual access to basic services (water, electricity and sanitation). Around 80-90% of the
households covered in our sample had pucca structures houses but only a small share had individual water connection, toilets and
bathrooms. Changes in housing surface and structure seems to be more relevant than changes in tenure status both of which, as will
be proved in the next subsection, can be generators of mobility.
Table 2: Changes in housing attributes for treated group following SRS
Attribute Before After
Construction
materials
80-90% pucca Pucca house
Surface 40% with areas > 225 sq.ft 225 sq.ft
Structure 14.9% had mezzanine or G+1 No vertical divisions
provided
Tenure 96.9% said to be structure
owners and only 2.6% said to be
on rental basis
Ownership papers,
transactions are
forbidden for 10 years
Access to
sanitation
94.7% didn‘t have bathroom
inside and 99.6% didn‘t have
individual toilets
Provision of separate
toilet and separate
bathroom
Access to water Only 31.6% had individual
water connection
Legal and independent
connection to water
Electricity
connection
97.6% had independent
electricity connection
Legal and individual
meters for electricity
connection
Settlement
structure
Village-type 3-5 multi-storey
buildings Source: Author‟s survey (2009)
4.3 Changes in budget constraint
Budget constraint equation defines what quantity of goods a household can buy based on goods prices and household salary. Lets
suppose a very simple equation where households utility is a function of a composite non-housing good (Nh) and a housing good (H).
If market equilibrium before rehabilitation is assumed, slums dwellers chose their previous housing maximizing their utility under
their budget constraint.
Vt = f(Nht, Ht) ; where Vt is households utility in time t, Nht corresponds to the non-housing good and Ht to housing good at
time t.
Housing good is composed of a series of specific attributes: (Bt) building structure, (Bst) provision of basic services and (St)
surface. Household‘s budget constraint is the following, assuming unitary price for non-housing good:
Wt= Nht +pBtBt + pBstBst + pStSt ; where Wt is households revenue and pBt, pBst, pSt are attributes prices.
If one holds non housing good constant, utility is considered to grow with an improvement in housing attributes. Therefore in-situ
slum rehabilitation is supposed to increase household‘s welfare. This might not be the case for slum resettlement policies since the
change of location inside the city might reduce household‘s utility. However changes in consumption of housing attributes generated
by the policy also affect the budget constraint equation. For instance, accessing legal basic service automatically generates changes in
the right side of the budget constraint equation since unitary prices are adjusted to existent tariffs. At the same time, the SRS might
also generate changes in the left side of the equation (revenue). The reallocation of time, previously used to collect water or to protect
the house, to income generating activities has been evidenced in studies carried out by Field (2003) in Peru and Aiga and Umenai
(2002) in Manila slums. Results from the authors survey also point in the direction of increased income through reallocation of time
International Conference On Applied Economics – ICOAE 2010 645
when control and treated groups are compared, suggesting that slum rehabilitation might be welfare improving. However no
statistical analysis has been carried out so far to prove the significance of policy effect on income.
A comparison of household's consumption expenditure as percentage of income between control and treated groups can be seen
in Figure 1. The rehabilitated (treated) group has considerably higher expenses in house maintenance, electricity and society charges
than the slum (control) group. Society charges correspond to expenses paid to cooperative society in charge of the rehabilitation
apartments which are normally used for maintenance of common spaces (corridors, lifts, etc). All of these items correspond to
changes in attributes due to the SRS policy and associated with formal housing. It is interesting to notice that expenses in water
services and solid waste collection remain about the same, however no information on quantity of water consumed was collected
during the survey so no further conclusions can be made. According to observations made during field surveys, slum pockets were
already served by Municipal Corporation which might explain little changes in water expenditure. In the case of waste disposal,
evidence suggest changes in households practices, as the proportion of households throwing waste in nearby trash cans augmented
from 75% before rehabilitation to 88% after wards and the proportion of households paying others to collect waste diminished from
13% to 8%.
Figure 1: Changes in consumer’s expenditure
Source: Author‟s survey (2009)
In the case of electricity, author's estimations suggest that control households have a higher per capita consumption (57,02
kWh/month.person) than treated households (48,09 kWh/month.person). This could be a consequence of the formalization of the
electricity supply. Sharma (2008) results for the case of MUTP resettlement indicate that 77% of the families say to be paying more
electricity than before. One of the highest increases in expenditure is the amount paid for maintenance and society charges, according
to Bhide et al. (2003) 67.3% of the respondents found this increase to be sharp and unaffordable. Sharp increases in electricity and
household maintenance are partly compensated by lower expenditure in health, education, food and transport. While transport and
education expenditures may reveal specific of the households employment and the slum pocket location, health expenditure might
actually reveal a positive indirect outcome of rehabilitation. A comparison between control and treated households from our survey
shows that on average rehabilitated households work more days a week than control households and lose less days of work due to
illness.
Households were also asked if they had problems facing end of months and shortfall frequency. According to results, around 25%
of treated households and 39% of control household face shortfalls regularly. Capacity of households to save money was measured in
the same manner. A higher proportion of households saving was found in the treated group and frequency of saving was also found to
be higher. In general terms, the rehabilitation of slum dwellers in formal housing generates both changes in household‘s expenditure
due to associated costs like having legal electricity, but also seems to have effects on households income through reallocation of time.
It is the magnitude of each of the effects that will define if households can maintain previous expenditures in new living conditions or
decide to move due to unsustainable and unaffordable new expenditures.
4.4 The opportunity cost
Literature refers to slum policies generated mobility, mostly as a consequence of the cost of opportunity. When a slum dweller
gets a formal ownership right, property value raises and so does the cost of opportunity. Payne (2001) arguments that the provision of
land titles may increase property values and displace most vulnerable groups in favor of higher incomes. Gilbert and Varley (1991)
explain how informal settlement that were once in peripheral or marginal location can now be found to be in strategic location given
urban growth.
In each of the sites, households were asked to estimate the rental and total value of their houses. Control groups which still live in
slums were asked about actual (slum) and future (rehabilitation) values while treated groups were asked about their previous (slum)
and actual (rehabilitation) values. Since property value is not only a function of housing attributes but of location, groups were not
merged and results are shown by site. As shown in Table 3, in all of the sites households estimate that the value of their houses
doubles following rehabilitation. In absolute values, households that are still living in the slums tend to overestimate the value of
rehabilitation tenements but in general terms the slum rehabilitation policy in Mumbai signifies a considerable transfer of physical
capital to slum dwellers.
Table 3: Housing valuation
Group Slum/rehab site name Sell (Rs.) Rent (Rs. Month)
Slum Rehabilitation Slum Rehabilitation
Treated Sai Wadi 625 000 1 532 268 2 185 7 769
646 International Conference On Applied Economics – ICOAE 2010
Treated Sundar Nagar 635 000 930 000 1 279 3 313
Treated Ganesh Nagar 581 539 1 509 700 1 327 4 250
Treated Ashram Chawl 570 833 1 253 036 2 325 4 622
Control Korba Mithagar 903 380 1 773 295 2 219 5 964
Control Sundar Nagar II 788 842 1 766 667 3 698 5 079
Control Betwala Chawl 603 000 1 800 000 3 100 17 250
Control Godiwala compound 663 333 1 600 000 4 143 7 000
Control Waterfield Road 1 474 429 3 000 000 6 000 7 750 Source: Author‟s survey (2009)
4.5 The reasons for moving out
Previous discussions suggest that while changes in housing attributes are mostly for the better this usually come at a higher cost.
If one plays with logic most of slum dwellers chose to live in slums either because they had an access problem - formal housing was
unaffordable - or because they had a durability issue - unable to support the cost of living in it. Under this approach, slum households
that have sufficient incomes to support new cost of living will probably be able to maintain sustainable trajectories. On the contrary,
for households having both an access and a durability issue, sustainable trajectories post-rehabilitation will be possible only if the
increase in income compensates the increased costs. When this is not the case households will quickly find themselves in a trap and
will probably return to slums.
During the author's survey reasons for moving out of original occupants were asked to new comers. Figure 2 reveals incompatible
living space and high cost of living as the two main reasons of mobility. Our results corroborate those found by Bhide et al (2003)
who asked all of surveyed households (new comers and original occupants) what they thought were the main reasons for people
moving out of rehab apartments. Although many were not able to respond, 27.5% of respondents thought that moving out was
primarily due to the weak economic status of slum dwellers and the increased maintenance charges. Insufficient living area was also
reported as a possible reason for mobility. While the first reason (unaffordable livelihood) reveals an alarming truth of how ―normal
standards‖ of living are simply not affordable for the poor, the second reason (incompatible living space) evidences a failure in the
policy to meet household needs.
Figure 2: Reasons for mobility of original occupants by new comers tenure status
Source: Author‟s survey (2009)
In fact, it seems that the standardization of the SRS policy - providing equal surfaces and housing structures regardless of original
conditions - might be its Achilles heel. Local authority‘s argument that giving areas of 225 sq.ft implies an improvement for most of
the slum dwellers but as the Slum Rehabilitation Society states four walls do not make a home. The rigidity of new houses poses a
serious problem since modification of housing attributes is restrained. Differences in type of investments in housing done between
control and treated groups evidence these limitations as 18.2% of Control groups investments in housing are used for structural
changes (adding divisions, mezzanine or doors) while only 8.6% in the case of the treated group. Results also suggest that a good part
of households investments in housing after rehabilitation are directed towards the acquisition of durable goods (electrical appliances
and furniture).
The latter point how sometimes slum dwellings, despite their constant stigmatization, seem to provide more adaptable
environments to household needs.Joshi (2006) states how ―The priorities of the slum dweller are frequently not those of the
authorities or the developers. Space takes precedence over permanence, function over aesthetic. A porch may be built before a
bathroom‖.
4.6 Recycling of poverty
Previous findings suggest that mobility post-rehabilitation is in many cases associated to a recycling of poverty. The poorest of
slum dwellers move to rehabilitation apartments and find themselves unable to support higher cost of living and are forced to move
out to slums. But what is the magnitude of this recycling and is it seriously damaging the overall efficiency of the policy? In order to
answer both of these questions it is necessary to know both original occupants destinations and new comers provenance. Net slum
absorption is given by the total of rehabilitated households minus poverty recycled households plus new comers coming for slums,
or:
International Conference On Applied Economics – ICOAE 2010 647
Net Slum Absorption = Rehab.households - Rehab.households went back to slums +
New comers that previously lived in slums
In the author's survey new comers were asked about their previous place of living and about original occupants destination.
Figure 3 shows how 43.5% of new comers were previously living in slums and that most of the original occupants moving out are
living in formal housing. Out of all the households leaving 18.2% went back to live in slums. Curiously it seems that mobility post-
rehabilitation is actually increasing the net absorption rates by allowing additional slum dwellers to enter formality. New comers
were also asked the time they spent in their previous residence and results indicate that around 72.7% had only spent 5 years or less
in their previous houses. This means that most of slum new comers are not eligible as beneficiaries given policy deadline (1st January
1995).
Figure 3: Net absorption rate
Source: Author‟s survey (2009)
Interestingly, actual status of tenure differs according to new comers origins. As 72.7% of those previously living in slums are
now tenants and 90.9% of those previously living in formal housing are now owners. In the case of original occupants 36.4% of those
renting their apartments moved back to slums and 45.5% to formal housing while 100% of those selling their houses moved to formal
housing (either in the city or outside of it). Both of which can be associated to the level of income of the household, as new comers
on rental status are on average poorer than new comers who bought rehabilitation tenements. The following table shows a
comparison of general indicators of control, original occupants (treated) and new comers groups. Results show how New comers are
very similar to original occupants but seem to belong to a slightly richer income group. A comparison between control and treated
groups evidences a diminution of household size and positive displacement of the income distribution post-rehabilitation. While the
proportion of children with less than 5 years suggests a diminution of fertility post-rehabilitation, smaller household size can also be
associated with individual ―invisible mobility‖. To cover for this, all treated households were asked if they had lost members after
rehabilitation. Results indicate that only 3.5% of treated households had lost members following rehabilitation and most of them left
the household due to marriage. The latter contradicts the hypothesis of ―invisible mobility‖ and suggests that most of the reduction in
household size is explained by diminished fertility.
Table 4: Original occupants & New comers
Control Treated –Original
Occupants
New comers
Household size 5.23 4.89 4.30
Proportion of children < 19 years 1.33 1.07 1.36
Proportion of children < 5 years 0.33 0.22 0.17
No. earners 1.71 1.58 1.87
% Permanently employed 28.7 33.1 39.4
Monthly income (Rs.) - % households in range
0 -2500 5.4 1.7 0.0
2501-5000 26.0 10.9 9.1
5001-7500 20.7 21.8 9.1
7501-10000 26.0 28.2 18.2
10001-12500 11.6 23.6 45.5
More than 12500 10.9 14.1 18.2 Source: Author‟s survey (2009)
5 Conclusions
This paper has presented new evidence on how slum policies impact slum dwellers livelihoods using residential mobility as an
indicator. There are three major findings. First, the household survey - as well as the two other surveys available - confirmed the
hypothesis of policy induced residential mobility. Abandonment rates after two and a half year of rehabilitation are around 10%
which is considerable when compared to the insignificant rates of residential mobility before the policy takes place.
648 International Conference On Applied Economics – ICOAE 2010
Second, the post-rehabilitation mobility analysis revealed that in some cases the benefits of rehabilitation are outstripped by
additional costs associated to new living status. In the case of in-situ rehabilitation, mobility was found to be associated to two
factors: a mismatch between household needs and new housing attributes; and an incompatibility between the high cost of living and
household‘s economic status. Results indicate that the first factor dominates since most of the original occupants moved out to formal
housing and not back to the slums.
Third, post-rehabilitation residential mobility was found to be linked to higher slum absorption rates. While this is certainly a
good thing, the generalization of this result is not automatic. Slum resettlement policies, in which slums are relocated at distant sites
might have higher proportions of residential mobility associated to poverty recycling due to greater impacts on the right side of the
budget constraint equation (ie. transport) and negative impacts on the left side of the equation (ie. loss of employment). Furthermore,
slums selected by the author's had achieved a certain degree of consolidation, and slum rehabilitation in less consolidated slums
might conduce to higher rates of poverty recycling.
Finally, throughout this paper it has been have proved how the analysis of residential mobility and abandonment destinations can
be used not as an indicator of sustainable policy trajectories but as an indicator of the mechanisms of slum formation in a given city.
Slum choice, according to results, is not only linked to formal entry barriers (accessing formal housing) but to a durability problem
(cost of living in formal housing).
6 References
Aiga,H. – Umenai,T. (2002), ‗Impact of improvement of water supply on household economy in a squatter area of Manila‘ Social
Science & Medicine, 55(4):627-641
Barry,M. (2006), ‗Formalising informal land rights: The case of Marconi Beam to Joe Slovo Park‘. Habitat international,
30(3):628-644
Bhide,A. – Shajahan,P.K. – Shinde,S. (2003), ‗Utilisation of slum rehabilitation scheme - a study conducted for the slum
rehabilitation authority, mumbai. Technical report, Tata Institue of Social Sciences
Cattaneo,M.D. - Galiani,S. - Gertler, P.J. - Martinez,S. and Titiunik,R. (2009), ‗Housing, health, and happiness‘. American
Economic Journal: Economic Policy, 1(1):75-105
Chandy, M. (2007), ‗Tdr- lessons from mumbai‘. In World Bank Urban Research Forum, 2007.
De Soto, H. (1990), ‗The other path: The invisible revolution in the Third World‘ bepress
Edwards,M. (1983), ‗Residential mobility in a changing housing market: the case of Bucaramanga, Colombia‘. Urban Studies,
20(2):131
Field,E. (2003), ‗Entitled to Work: Urban Property Rights and Labor Supply in Peru‘. Quarterly Journal of
Economics,122(4):1561-1602
Field,E. (2003), ‗Fertility responses to urban land titling programs: The roles of ownership security and the distribution of
household assets‘. unpublished paper, Harvard University, 2003.
Gilbert,A. – Varley,A. (1991), ‗Landlord and tenant: housing the poor in urban Mexico‘
Gilbert,A. – Ward,P.M. (1982), ‗Residential movement among the poor: the constraints on housing choice in Latin American
cities‘ Transactions of the Institute of British Geographers, 7(2):129-149
Gilbert,A. (1999), ‗A home is for ever? Residential mobility and homeownership in self-help settlements‘. Environment and
Planning A, 31:1073-1092
Gulyani,S. – Talukdar,D. (2008), ‗Slum real estate the low-quality high-price puzzle in nairobi's slum rental market and its
implications for theory and practice‘. World Development, (10)
Hagn,A. (2006), ‗Looking at the situation of informal settlements in mumbai‘ Mumbai Reader 06 UDRI
Jacobsen,I. (2003), Privatisation and individualisation of land rights in urban areas: The cases of Imizamo Yethu and Joe Slovo
Park in Cape Town‘. PhD thesis
Joshi,P. (2006), ‗Slums a solution to the housing problem‘ Mumbai Reader 06 UDRI
Lall,S.V. - Suri,A. and Deichmann,U. (2006), ‗Household savings and residential mobility in informal settlements in Bhopal,
India‘. Urban Studies, 43(7):1025
McFadden,D. et al. (1978), ‗Modelling the choice of residential location‘. Spatial interaction theory and planning models, 25:75-
96
MCGM (2005), ‗Mumbai city development plan 2005-2025‘. Technical report
Montgomery Watson and Yuva (2001). ‗Slum sanitation project‘. Technical report
Mukhija,V. (2001), ‗Enabling slum redevelopment in Mumbai: policy paradox in practice‘. Housing Studies, 16(6):791-806
Payne,G. (1977), ‗Urban housing in the Third World‘. Routledge Kegan & Paul
Payne,G. (2001), ‗Urban land tenure policy options: Titles or rights?‘ Habitat International, 25(3):415-429
Sharma,R.N. (2008) ‗Impact asessment of ressettlement implementation under the MUTP program. Technical report, Tata
Instutute of Social Sciences for The Mumbai Metropolitan Region Development Authority (MMRDA)
Turner,J.C. (1968), ‗Housing priorities, settlement patterns, and urban development in modernizing countries‘. Journal of the
American Planning Association, 34(6):354-363,
United Nations (2007), ‗World urbanization prospects: The 2007 revision‘
Top Related